Esophageal adenocarcinoma (EAC) incidence is increasing rapidly and endoscopy-based screening has not reduced mortality despite targeting patients with known risk factors [GERD, obesity, smoking and Barrett's esophagus (BE)]. This proposal describes a five-year training and research plan to develop expertise in the control and prevention of EAC through a K07 career-development award. The long-term research goal is to develop clinical prediction models that enable timely utilization of health resources in the detection of curable EAC. These models would incorporate serum and tissue biomarkers, known risk factors and modifiable patient-medical system interactions, such as barriers to patients presenting for and receiving health care, which might facilitate detection of curable EAC. This proposal has important public health implications in that it seeks to improve risk stratification for patients with EAC by examining deficits in understanding of the patient-medical system interaction while determining differences in underlying biology between EAC with associated BE compared to those without associated BE. The University of Pittsburgh is the ideal venue for this work, offering unparalleled research support, high patient volumes and world-renowned mentorship in EAC biomarkers, epidemiology, screening (Luketich, Vaughan and Jobe), and measurement development (Switzer). The long-term career goal to develop clinical prediction models for EAC will be initiated with the short-term, integrated research and training plan described in this K07 proposal. A novel self-report instrument for use in a cross-sectional study of EAC will be developed to examine the hypothesis that patients presenting with advanced versus early-stage EAC will have a different prevalence and pattern of risk factors (Aim 1) and less effective interactions with the medical system (Aim 2; e.g. lack of regular primary care provider, self- versus physician-directed treatment of GERD). Using the current symptom-based screening paradigm, however, EAC may not be detectable at a curable stage in a large subset of patients if the cancer does not arise from endoscopically visible BE. To address this question, proteomic expression patterns will be determined (Aim 3) between EAC with associated Barrett's esophagus compared to those without associated Barrett's with respect to location (esophageal body versus gastroesophageal junction) and stage. Coursework is proposed that builds upon prior training. Specific expertise in cancer biomarkers, measurement development, and prospective study design, implementation and analysis will be obtained. With strong institutional support and mentoring, the candidate is perfectly positioned to complete the proposed training and research. The findings will provide preliminary data for a R01 to be submitted in the fourth year of the award period. This application has specific relevance to the primary mission of the National Cancer Institute, which is to support education and training in fundamental sciences and clinical disciplines, establishing the groundwork for this candidate's future success as an independently funded researcher in the prevention and control of esophageal cancer.